Certainly Aristotle, who puts forward the orthodox Peripatetic theory of the heavenly bodies (928 E ff.), is only a name chosen by Plutarch to signify the school that he represents (cf. 920 F), even as the representative Epicurean in is called Epicurus.There is no reason to change Ἐπίκουρος of the MSS. in 548 A to Ἐπικούρειος, as Fabricius did. Aristotle here supports Epicurus there.

The Stoic position is represented by Pharnaces. This name was borne by the son of Mithridates, of whom Plutarch tells in the Lives of Pompey and Caesar, as well as by several notable Persians mentioned by Herodotus and ThucydidesThere was also a city in Pontus named Pharnaceia (, 17 [502 F]).; and Plutarch probably chose it for his Stoic because of its Asiatic sound.Hirzel (Der Dialog, ii, p. 186, n. 4) says that Pharnaces is certainly a former slave, one who had shared the fate and sentiments of Epictetus. This, of course, is the merest fancy; not all Asiatics, not even all in Rome at this time, had been slaves. For Athenians named Pharnaces cf. ii². 1039. 84 and 202. 55.

After the role of Lamprias the largest in the dialogue proper is that of Lucius, who is probably the same as Lucius, the pupil of Moderatus the Pythagorean, from Etruria, a guest at the dinner which Sulla gave for Plutarch in Rome ( viii. 7-8 [727 B ff., 728 D ff.]).Another Lucius, the son of Florus, appears in vii. 4 [702 F]; cf. Ziegler, , 55. Early in the dialogue (921 F) Lamprias turns to Lucius for aid; he seems to think it appropriate that Lucius should set forth the strict demonstration of the Academic theory concerning the moon (cf. 928 D - E); and in fact the statement and defence of this position are shared by the two of them.It is Lucius who demands that the Stoic theory should not be passed over without refutation (921 F). It is he who replies when Pharnaces complains of Lamprias’s violent treatment of the Stoics (922 F). His speeches extend from 922 F to 923 F, where Lamprias takes over to give him time to collect his thoughts, from 928 F to 929 E, from 930 A to 931 C, and from 931 D to 933 E.

Theon, whom Lamprias asks to identify a quotation (923 F) and whom he later teases for admiring Aristarchus to the neglect of Crates (938 D), is recognized as the literary authority in the group (cf. 931 E, 940 A). He is probably to be identified with Θέων ὁ γραμμοτικός, who was a guest at Sulla’s dinner along with Lucius ( 728 F) and who also dined with Plutarch at the house of Mestrius Florus ( 626 E).This Theon, whose home was Egypt (cf. 939 C - D), is certainly not the same as Θέων ὁ ἑταῖρος ( 620 A, , 386 D), who is probably the Theon of , and 667 A and 726 A ff. In the his chief contribution is a speech (937 D 938 C) which he makes after the main part of the dialogue has been concluded and which Lamprias praises as a kind of relaxation after the seriousness of the scientific discussion.

The last of the persons present is Menelaus the mathematician. Lucius addresses him directly once (930 A), but Menelaus makes no reply and neither speaks nor is spoken to elsewhere in the dialogue as we have it,Unless the plural ὑμῖν used twice by Lamprias at 939 C - D is meant to include Menelaus as well as Theon; cf. note a on p. 170 s.v.. He is not mentioned anywhere else by Plutarch either; but he is probably meant to be the Menelaus of Alexandria whom Ptolemy once calls ὁ γεωμέτρης and twice cites for astronomical observations which he made at Rome in the first year of Trajan (A.D. 98).Ptolemy, , vii. 3 (ii, p. 30. 18 ff. and p. 33. 3 ff. ([Heiberg]); cf. Orinsky, s.n. Menelaos 16 in Pauly-Wissosa, xv. 1. 834 f.

3. From 937 C - D it follows that the interlocutors have hitherto been promenading as they talked and that now they sit down upon the steps, seats, or benches (ἐπὶ τῶν βάθρων) and remain seated to the end. No other indication of the scene or location is given in the work as we have it. It had generally been assumed that the dialogue was meant to take place in Chaeroneacf. Hirzel, , ii, p. 184, n. 1, who discusses and rejects the suggestion that the scene is Delphi. Raingeard in his note on 939 C (p. 129 of his commentary) says that ὥσπερ ἄνω περὶ Θήβας there would allow the inference that the speakers are on the coast of Egypt. No such inference is justified by this phrase, of course; in fact, the preceding ὕλην δὲ καὶ καρποὺς αὐτοῦ (or αὐτόθι, as Raingeard conjectures) μὲν ὄμβροι τρέφουσιν and the subsequent παρ’ ὑμῖν ἐν Αἰγύπτῳ (939 D 1) show that the scene of the dialogue is not anywhere in Egypt.; but nothing in the text requires this, and F. H. Sandbach has adduced strong arguments for believing that the dramatic location is Rome or the vicinity of Rome.F. H. Sandbach, The Date of the Eclipse in Plutarch’s De Facie xxiii (1929), pp. 15-16; cf. Ziegler, , 73-74. I am indebted to Mr. Sandbach for sending me, along with copies of his publications, many of his unpublished opinions concerning points in the De Facie and copies of his correspondence with J. K. Fotheringham occasioned by the publication of the article cited above. The persons in the dialogue furnish one of these arguments. Apollonides, Aristotle, and Pharnaces occur nowhere else in Plutarch’s writings and are probably all fictitious characters. Plutarch nowhere else mentions Menelaus the mathematician either, but we know that Menelaus spent some time in Rome (see note a, p. 8). Sulla, Lucius, and Theon all appear together at a dinner given for Plutarch when he had returned to Rome after an interval of absence ( viii. 7-8); and none of these three is ever mentioned as being anywhere but in Rome or its vicinity (see § 2, supra). Lamprias alone belongs to Plutarch’s circle at Chaeronea; but it is by no means certain that he did not visit Rome as Plutarch did, though there seems to be no definite evidence either way.Lamprias at least pretends to be conversant with Latin ( 726 E ff.). On Plutarch’s visits to Rome cf. Ziegler, , 19-20.

The other argument for the dramatic location is connected with the question of the dramatic date of the dialogue. At 931 d-e Lucius refers to a recent total solar eclipse, saying: if you will call to mind this conjunction recently which, beginning just after noonday, made many stars shine out from many parts of the sky δότε μοι, ταύτης ἔναγχος τῆς συνόδου μνησθέντες ἥ πολλὰ μὲν ἄστρα πολλαχόθεν τοῦ οὐρανοῦ διέφηνεν εὐθὺς ἐκ μεσημβρίας ἀρκαμένη GinzelF. K. Ginzel, (Berlin, 1899), pp. 202-204; cf. also Plates X and XI for the paths of solar eclipses during the first and second centuries a.d. The data for the eclipses of 75 and 83, s.v., come from Ginzel’s tables, Op. cit. p. 78 and pp. 110-111. identified this eclipse with the one which occurred on 20 March a.d. 71, for he found that all other solar eclipses visible in Chaeronea during Plutarch’s lifetime fell too far short of totality to permit the appearance of stars. His conclusion was generally acceptedStruyck (cited by Ginzel, Op. cit. p. 203) appears to have come to this conclusion before Ginzel; and Ginzel’s identification was accepted by M. Adler ( [Nikolsburg, 1910], p. 4) and by Fotheringham as cited by A. O. Prickard ( [1911], p. 75, and , ii, p. 253). Hirzel (, ii, p. 182, n. 1), following Volkmann, does not even mention the eclipses of 59, 71, and 75, which Ginzel held to be the only ones worthy of consideration. until SandbachOp. cit. in note c, p. 8 supra. pointed out that, since this eclipse reached its maximum phase at about 11 A.M. local solar time in Chaeronea,10hr, 58m, 4 according to Ginzel (Op. cit. p. 204); 11hr, 4m, 1 according to Fotheringham as quoted by Prickard (, ii, p. 253). Plutarch could not have referred to it as having begun after noonday. Ginzel had assumed that the place of observation was Chaeronea; Sandbach, having recognized that this assumption is unwarranted, was able to consider two other eclipses, that of 5 January A.D. 75 and that of 27 December A.D. 83. The latter was total at Alexandria shortly before 15 hours. The former was total in Carthage a little after 15 hours and in the latitude of Rome on the eastern side of the Adriatic at about 15 hours, 20 minutes; at Rome itself the maximum obscuration was 11 · 5 digits, so that, since according to Fotheringham (1921), cited by Fotheringham in a letter to Sandbach (22 January 1929); in this letter Fotheringham states that a certain number of stars were visible at Rome in 75. cf. Ginzel, Op. cit. p. 14: Bei den zentralen Sonnenfinsternissen einzelne Sterne treten mitunter hervor, bevor die Phase 11 zöllig geworden ist. stars other than Venus have been visible where the solar obscuration was 10 · 7 digits, it is perfectly possible that some stars would have been seen at Rome about 3.20 p.m. local solar time on 5 January A.D. 75. This eclipse of A.D. 75 as seen in Rome certainly fits the conditions of Lucius’ statement better than does the one of A.D. 71 as seen in Chaeronea, even though it was rather late to be described as beginning just after noonday.Its beginning, which would have been at approximately 13.50 hours, could not have been observed with the naked eye; but Plutarch was capable of calculating it roughly. In any case, whether the συνόδον ἣ ἀρξαμένη is to be taken strictly or in the sense of the time when darkness began, μεσημβρία, as Sandbach has said, is an extended period of time and not an astronomical moment; and Lucius means that the conjunction began just after noonday was over. It must be emphasized that there is no reason to assume that Plutarch himself saw the eclipse to which Lucius refers. He had undoubtedly heard that it had been seen in or near Rome; he almost certainly had seen the eclipse of A.D. 71 in Chaeronea and may have seen that of A.D. 83 in AlexandriaWe do not know when Plutarch visited Alexandria. In v. 5 (678 C ff.) his grandfather is present at a banquet given for him after his return from Alexandria. Sandbach (loc. cit.) thinks that this could have been after 83; but, whether this is so or not, we do not know whether there may not have been more visits to Alexandria than this one.; and what he had seen during one or both of these eclipses he may very well have applied to the eclipse of A.D. 75, which he had not seen.If 932 B (περιφαίνεταί τις αὐγὴ περὶ, τὴν ἴτυν . . .) means, as has sometimes been supposed, that Plutarch had seen the corona, he must have had this experience in 71 or 83. No one in or near Rome would have seen it in 75. I doubt that these words apply to the corona at all, however, for the subsequent oὐκ εὦσα βαθεῖαν γενέσθαι τὴν σκιὰν καὶ ἄκρατον would be a remarkably tame way of describing that spectacle. If the passage refers to any observed phenomenon, it is more likely to have reference to an annular eclipse. We may then conclude that the dramatic date of the dialogue is later than A.D. 75, but how much later it is remains uncertain despite Lucius’ reference to the eclipse as recent. The word which he uses, ἔναγχος, like the English recent, has a meaning relative to its context, and in the case of anything so unusual as a total solar eclipse might refer to an event that had taken place at any time within a decade or more; it seems in this passage not to be used of the immediate past, for Lucius expressly reckons with the possibility that his audience may not recall the recent conjunction and may have to fall back upon literary evidence for the impression made by a total solar eclipse.931 E: εἰ δὲ μή, Θέων ἡμῖν τὸν Μίμνερμον ἐπάσει, κτλ. Of course, this is primarily a literary device to excuse the introduction of the literary references; but it shows that Plutarch does not expect his readers to remember what a total solar eclipse is like. The attempts to find a historical reference in 945 B which would help to fix the date of the dialogue are quite perverseHirzel (, ii, p. 182, n. 1) excised Τυφών in 945 B (Τιτυοὶ δὲ καὶ Τυφῶνες ὅ τε Δελφοὺς κατασχὼν καὶ συνταράξας τὸ χρηστήριον ὕβρει καὶ βίαλ Τυφὼν ἐξ ἐκείνων κτλ.), took ὁ συνταράξας βίαλ as a reference to Nero, and concluded that Plutarch must have written this after the devastation of Delphi and before the restoration of the oracle. Adler (, etc. [see note a, p. 10], pp. 5-7) defended the text of the mss., which he interpreted to mean demons of the nature of Tityus and Typho and among these especially the Typhon who, etc., and followed Pomtow ( li [1896], pp. 377 ff.), who showed that the extinction of the Delphic oracle during the time from Nero to Hadrian was pure invention and who took Τυφών in , 945 B, as a reference to the conflagration in 83 B.C. Adler then, assuming that after the ceremonious restoration of the temple in A.D. 84 Plutarch would not remind his readers of its devastation, concluded that the dialogue must have been written before A.D. 84. This argument was criticized by K. Mras (, lxv [1914], p. 187), who in turn deleted Τυφὦνες from the text and read Τιτυοὶ δὲ καὶ ὁ Τυφὼν ὁ νφών Δελφούς βίαλ κτλ. This violent alteration is even less justifiable than Hirzel’s excision of Τυφών, with which it shares the fault of producing the hiatus βιάλ ἐξ; but the text of the MSS. is impossible despite Adler, for (a) one does not say in any language such creatures as Tityus and Typho and in particular Typho . . ., (b) nowhere else is Typho himself said to have done the deed here ascribed to him, and (c) a reference to the conflagration is at least as improbable as the supposed reference to Nero. Kaltwasser’s change of Τυφών to Πύθων, on the other hand, is practically certain. Confusion of π and τ and of θ and φ is easy and common, and πύθων coming after τυφώνες would very easily be assimilated to it. Moreover, in , 421 C, τὰ περὶ Πύθωνα are included among δαιμόνων πάθη along with τὰ Τυφωνικά and τὰ Τιτανικά. In 414 A - B the oracle at Delphi is said to have been long deserted in what is represented as ancient times; and, if it is denied that the beast (which is not here named but is certainly Pythons3) was the cause, that is done in order to ascribe the cause to Δαίμονες. Finally, Πύθων and Τιτυός are named together by Plutarch in Pelopidas, 16 (286 C) as they are by Strabo (ix. 3. 12 [cc. 422-423]) and Apollodorus (, i. 4. 1. 3-5 [22-23]).; and we are restricted by the evidence at present available to the conclusion that the conversation was meant to have taken place in or about Rome some time and perhaps quite a long time after A.D. 75.

So much for the dramatic date. There is no reason at all for Hirzel’s assertion, ii, p. 184, n. 1. that this and the date of composition coincide. Certain striking similarities between the and the have often been observed, but from these can be drawn equally cogent — and equally hypothetical — arguments for the priority of either to the otherM. Adler ( x, pp. 115-116) contends that in the Plutarch excerpts the ; but see Raingeard, p. xxviii of his edition of .; and, since in any case the date of the is uncertain,Ziegler, , 76, gives about 100 as the date; but cf. Flaceliére, , note 4 and pp. 13-17. the relative chronology of the two if established would not determine the date of the .

4. The structure of the De Facie is complicated. The whole of the work is narrated by Lamprias who speaks in the first person and quotes those who took part in the conversation, including himself, some few times in indirect discourse (e.g. 933 F) but for the most part directly. The last part of his narration (chaps. 26-30 [940 F 945 D] consists entirely of Sulla’s myth given in Sulla’s own words; this myth, Sulla himself says, is a story told to him by an unnamed stranger, whom he quotes first indirectly and then (942 D ff.) directly to the end. The second or eschatological part of the myth the stranger had told Sulla that he had himself heard from the chamberlains and servitors of Cronus (cf. 945 D). Hearing it from Lamprias now, the reader has this part at fourth hand and the geographical introduction of the stranger at third hand.Cf, Plato’s and Shorey, , p. 287.

From 937 C it appears that Sulla had promised to tell his myth in return for an account of what had been said in an earlier discussion about the nature of the face which appears in the moon. Such a compact may have been expressly made in the beginning of the dialogue which is lost, where Sulla may have come upon the company already engaged in reviewing that earlier discussion (see note a, p. 3). So much is no more than conjecture. It is certain, however, that most of what Lamprias narrates from chapter 2 through chapter 23 is a conversation which is itself represented as containing a resume or report of what was said at an earlier conversation. This the beginning of chapter 24 (937 C) states explicitly: ἡμεῖς μὲν οὖν, ἔφην, ὅσα μὴ διαπέφευγε τὴν μνήμην τῶν ἐκεῖ λεχθέντων ἀπηγγέλκαμεν, and the ἐδόκει λέγεσθαι at the end of chapter 2 (920 F) implies that what Lamprias has hitherto said in that chapter had been used as an argument in the earlier discussion. The leader of that discussion, which is referred to as a διατριβή,By Lucius at 929 B: ὁ μὲν οὖν ἑταῖρος ἐν τῷ διατριβῇ τοῦτο ἀποδεικνὺς ἠυδοκίμησεν. was not Lamprias or Lucius, who here recapitulate it,cf. besides 937 C, 920 F, and 929 B, which have already been cited, especially 921 F, 930 A, 932 D, 933 C. but someone to whom Lamprias, Lucius, and Sulla refer as our comrade and who probably is meant to be Plutarch himself.cf. 921 F, 929 B, 929 F, and see note a on p. 48 s.v.. Lamprias and Lucius are, of course, presumed to have been present at that discussion with their comrade and Sulla to have been absent from it.The logic of the situation demands this; but it is also implied by Sulla’s question at 929 F. Of the others, Apollonides certainly was not present,This is implied by his question in 920 F and confirmed by that in 921 B: ἀλλὰ πῇ τὸν ἔλεγχον αὐτῷ προσῆγες; (in this latter passage Pohlenz [ xxxii, 1912, p. 649] argued for retention of the MSS, reading, προσῆγε, understanding as subject ὁ ἑταῖρος, who he assumes was mentioned in the lost beginning of the dialogue; but surely this sentence is too far from even such a hypothetical antecedent, and Adler’s προσῆγες is an obvious and highly probable correction). nor was TheonThis is certainly implied by his interchange with Lucius in 932 D - E.; Pharnaces probably was notThis is the most reasonable inference to be drawn from 921 F, where Lucius requests that Pharnaces be given some consideration, and from Pharnaces’ comment in 922 F upon the attack of Lamprias. Nevertheless, Pharnaces’ words in the latter passage, ἐμὲ δ’ οὖν οὐκ ἐξλαξεσθε τήμερον κτλ., are open to the interpretation that he had been present at the earlier discussion and had there been drawn out by the Academic gambit.; and concerning Aristotle and Menelaus the text as we have it allows no clear inference to be drawn.Lucius’s one remark to Menelaus (930 A), αἰσχύνομαι σοῦ παρόντος κτλ., seems to imply that the latter had not been present at the earlier discussion; but this is not decisive, especially in view of the fact that Menelaus makes no reply. Aristotle’s silence when Lamprias addresses him in 920 F might be taken to mean that he had heard this before; and πρὸς Κλέαρχον, ὦ Ἀριστότελες, ἐδόκει λέθεσθαι τὸν ὑμέρτον could be interpreted as a reminder, although what follows, ὑμέτερος γὰρ ἁνὴρ κτλ., sounds as if this were something new. In 929 B Lucius in a speech addressed especially to Aristotle refers to what our comrade said ἐν τῇ διατριβῇ and adds that he will not repeat what he learned παρ ὑμῶν μεθ ὑμῶν, which might be taken to imply that Aristotle too had attended the διατριβή in question, although it might have a more general meaning. What these men other than Lamprias and Lucius say in chapters 2-23 is not, then, part of the report of that earlier discussion; but neither is all that Lucius says, for in several places his remarks or arguments are expressly declared to be his own contribution.cf. Lamprias’s comment, οὐχ οὕτως δ’ ὁ ἑταῖρος ἡμῶν, in 921 F and his καλῷ λόγῳ καλὴν ἀναλογίαν προσλεθηκας: οὐ γὰρ ἀποστερητόν σε τῶν ἰδίων (931 D). The latter marks the last sentence of Lucius’s preceding speech (δότε δή μοι γεωμετρικῶς εἰρεῖν κτλ.) as his own, while Lucius’s own subsequent statement (οὐκοῦν καὶ δεύτερον ἀναλογίᾳ προσχρητέον) makes the same claim for what follows. In 933 C (παρίημι δ’ ὅσα ἐλέχθη) and possibly in 929 B (ἐγὼ δὲ ταῦτα μὲν οὐκ ἐρῶ κτλ. [see note b supra]) Lucius indicates that he is not giving a full account of the earlier discussion. That earlier discussion cannot, however, be identified with any that Plutarch may have had with his friends or with any lecture that he may have given; it is primarily a literary fiction, part of the structure of the dialogue for which it provides a specious motivation.

The recapitulation of this fictitious discussion along with the incidental arguments provoked by it contains all that Plutarch would consider to be scientific in the dialogue. At its conclusion Lamprias is ready for Sulla’s myth (chap. 24 init. [937 C - D]); but before Sulla can begin to speak Theon raises the question of the habitability of the moon, contending that, if it is not habitable, there can be no reason for it to exist with the nature or composition that according to Lamprias and Lucius it does have.cf. 937 D: εἰ δυνατὸν ἐκεῖ κατοικεῖν. εἰ γὰρ δυνατόν, ἄλογον καὶ τὸ γῆν εἰναι τὴν σελήνην: δόκει γάρ πρὸς οὐδὲν ἀλλὰ μάτην γεγονέναι κτλ. Lamprias calls Theon’s speech a kind of relaxation after the seriousness of the preceding discussion. In fact, however, Theon has raised the metaphysical problem of the final cause; and to this Lamprias replies at length (chap. 25). He argues first that the moon, constituted as he contends it is, need not, even if uninhabitable, be without a purpose in the universe (938 C - F), and secondly that, even if uninhabitable by corporeal human beings, it may still be inhabited by living beings of an entirely different kind to whom the moon may justly appear to be the only real earth and our earth the slime and dregs of the universe, uninhabitable by creatures that have warmth and breath and motion. Here Sulla checks Lamprias (chap. 26 init. [940 F]) lest the latter encroach upon his myth; and Lamprias was upon the very threshold of it, for the myth, as it turns out, teaches that the moon is inhabited by souls that have left their bodies after death on earth or have not yet been incorporated by birth into terrestrial bodies. So the episode consisting of Theon’s speech and Lamprias’s reply (chaps. 24-25) is not merely a formal literary device. It is, to be sure, a transition from the scientific part of the dialogue, in which it is argued that the lunar phenomena imply the earth-like constitution of the moon, to the concluding myth in which the purpose of such a moon in the universe is imaginatively portrayed; but this transitional episode raises the philosophical question, without the answer to which the strictly astronomical conclusion could to a Platonist or Aristotelian be no complete or satisfactory explanation, and itself contains the metaphysical answer, of which the myth is, despite all its intrinsic interest, essentially a poetical embellishment. When this transition is properly attended to, there can be no question about the integral unity of the whole dialogue or any doubt that the purpose of the whole is to establish and defend the position that the moon is entirely earthy in its constitution and that on this hypothesis alone can the astronomical phenomena and the existence of the moon itself be accounted for.cf. M. Pohlenz, clxxx [1918], p. 323.

5. The main part of the dialogue is of extraordinary interest for the history of astronomy, cosmology, geography, and catoptrics; and this aspect of the work deserves more attention than it has usually received.J. O. Thomson, History of Ancient Geography (Cambridge, 1948), pp. 330 f., gives a brief outline of this part of the work and cites Duhem’s and Humboldt’s praise of it. A. O. Prickard has some general remarks on the subject in the introductions to his two translations of the dialogue ( [Winchester and London, 1911], pp. 9-15, and , ii [Oxford, 1918], pp. 246-253). So has S. Günther in his outline of the dialogue, (Braunschweig, 1911), pp. 24-35, nearly half of which, however, is concerned with the myth. Hirzel in his treatment of the dialogue ( [Leipzig, 1895], ii, pp. 182-189) has little or nothing specific to say of its scientific aspect. The most extensive monograph on the dialogue, Maximilian Adler’s ( x [1910], pp. 85-180), is concerned with the scientific passages only in so far as the author thinks that from them he can draw support for his thesis that Posidonius was Plutarch’s source for the dialogue. A similar purpose limits the treatment of the work by K. Praechter in his (Leipzig, 1901), p. 26 and pp. 109-120. cf. also the notes of W. Norlind, , xxv (1927), pp. 265-277. It is not a technical scientific treatise and is not to be judged as if it were meant to be such; but it is all the more significant that in a literary work intended for an educated but non-technical audience towards the end of the first century a.d. Hipparchus and Aristarchus of Samos are familiarly cited and a technical work of the latter is quoted verbatim, the laws of reflection are debated, the doctrine of natural motion to the universal centre is rejected, and stress is laid upon the cosmological importance of the velocity of heavenly bodies. It is interesting to compare the treatise of Ibn Al-Haitham (965-1039) which was translated from the Arabic by Carl Schoy under the title (Hannover, 1925). Ibn Al-Haitham’s explanation of the face is that the nature of the moon’s substance must differ from place to place, since the variation in illumination can be the result only of a difference in the power to absorb and reflect light, and the spots are places of greater density and less power of absorption (pp. 20 ff. and 29-31). Though Schoy appears to have been unaware of it and Plutarch does not mention it, this explanation is ascribed to οἱ ἀρὸ τῶν μαθηματικῶν in Aëtius, ii. 30. 7 (, p. 362. 5-13). Ibn Al-Haitham rejects the theory that the spots are shadows cast by prominences on the moon, arguing that such shadows would not always have the same shape and position, as the spots do (pp. 14-17). Like Plutarch, however, he knows and refutes the notion that they are a reflection of the terrestrial ocean or any other terrestrial feature (pp. 1-2, 5-7; , chaps. 3-4); and he also adduces the colour of the moon in eclipse (pp. 31 f.; , 934 B - D). He proves impossible as well (pp. 4-5, cf. p. 2) an explanation unmentioned by Plutarch but recorded by Simplicius (, p. 457. 25-30) that the spots are the result of vapours rising from below and obscuring the moon’s brilliance (cf., however, for something similar, Milton, , v. 415-420, and , 922 B - C). Like Cleomedes (ii. 4. 103 [p. 186. 14-27 Ziegler]), Ibn Al-Haitham seems to hold that the moon as a reflecting convex mirror would have to appear as a single point of light (pp. 7 f. with Schoy’s note, p. 8, n. 1). Most of the attention given to the dialogue, however, has been attracted by the concluding myth.It was probably the myth as much as the more strictly astronomical part of the dialogue that caused Kepler to make his Latin translation and commentary of the , which he did shortly before his death. This is printed in volume viii of , ed. Dr. Ch. Frisch (Francofurti a. M., 1870). cf. R. Schmertosch, Keppler zu Plutarchs Schrift Vom Gesicht im Monde, (Leipzig, 1897), pp. 52-55, and R. Pixis, (Munich, 1899). This consists of two parts. The second and main part is the eschatological myth, which establishes the purpose of the moon in the cosmos by explaining her role in the life-cycle of souls and which the stranger told Sulla he had from the chamberlains of Cronus (942 D 945 D); the first is the introduction to this myth or frame-story, in which the stranger explained to Sulla how from the continent on the other side of the Atlantic he came to the Isle of Cronus, one of several that lie westwards of Britain, and thence, after having served thirty years, travelled to Carthage where he met Sulla (941 A 942 C).

This geographical introduction has aroused the wildest speculations. Kepler was convinced that the trans-Atlantic continent was America, and he tried to identify the islands mentioned in the mythcf. notes 97, 98, 103, and 105 to Kepler’s translation (see note a, p. 20 supra) and note 2 to his . In (Antwerp, 1593), p. 5, this passage of Plutarch was used, apparently for the first time, to prove that the ancients knewthe American continent.; W. Christ in 1898 still could assert that Plutarch’s continent is obviously America and proves that about A.D. 100 sailors reached the North American coast via Iceland, Greenland, and Baffinland (1898), p. 662, n. 1. W. Schmid and O. Stählin in the sixth edition of this work (Zweiter Teil, Erste Hälfte [1920], p. 498) suppress this note of Christ’s but write aus dem Festland jenseits des atlantischen Ozeans (Amerika ?). ; and in 1909 G. Mair argued that the source of this knowledge of America was reports of Carthaginian seafarers who had penetrated into the Gulf of Mexico, that the Isle of Cronus is Scandinavia, and that the northern geography of the myth derives from accounts of the voyages of Pytheas of Massilia.G. Mair, Pytheas’ Tanais und die Insel des Kronos in Plutarchs Schrift Das Gesicht im Monde (, 1909). A fair example of Mair’s argument is his identification (p. 18) of the κόλρος mentioned in 941 B with the Christiana-Fjord, although according to Plutarch it is in the trans-Atlantic continent. Moreover, all of Plutarch’s islands lie to the West and North-West of Britains3 Even before Mair had published his fantastic theory Ebner had conclusively demonstrated that Plutarch could not have referred to any real crossing of the Atlantic or any rumours of such a crossing, that by using the name Ogygia at the beginning (941 A - B) he had clearly indicated the purely mythical intention of his geography, and that this geographical setting is simply an imitation of Plato’s Atlantis in the spirit of Hecataeus story of the Hyperboreans, Theopompus Meropis, and the Sacred Records of Euhemerus.E. Ebner, (Munich, 1906). A. von Humboldt had concluded long before that the geographical frame is entirely mythical ( [Berlin, 1836], pp. 174-185). H. von Arnim (Plutarch über Dämonen und Mystik, pp. 37-47 [, Afd. Letterk., 1921]) contended that Plutarch’s source for chapter 26 was a fantastic travel-romance that had nothing to do with philosophy or moon-demonology, but in which the demons of Cronus served the purpose of prophesying to the hero about his future. W. Hamilton ( xxviii [1934], pp. 24 ff., cf. p. 24, n. 1), while citing as parallels to the geographical myth Hecataeus, Euhemerus, Theopompus, and the Abaris of Heraclides Ponticus (cf. also Hirzel, , ii, p. 187, n. 4), maintains that Plutarch wrote the whole of his myth in direct imitation of Plato’s story of Atlantis. Rohde (, 204-276 = 3rd edition [Leipzig, 1914], pp. 219-296) places Plutarch’s geographical myth in its proper environment with the romances of Theopompus, Hecataeus, Euhemerus, Iambulus, Antonius Diogenes, and Marcellus. cf. also H. Martin, (Paris, 1841), i, pp. 290-304, and J. O. Thomson, Op. cit. (note b, p. 18), pp. 237-238. The additional geographical particulars are the usual corroborative detail intended to give artistic verisimilitude to an otherwise bald and unconvincing narrative. The theme of the sleeping Cronus may have been suggested to Plutarch by Demetrius of Tarsus, who in the (419 E 420 A) is made to say that on an island near Britain Cronus is kept prisoner by the bonds of sleep and is guarded by Briareus and attended by Spirits who are his servitors. This Demetrius appears to have been an historical person who did travel to Britain, whence in the dialogue he is said to have recently returned; and he may have told Plutarch some Celtic legend or superstition which the latter hellenized and wove into the fabric of his myth.For Demetrius cf. R. Flaceliére, (Paris, 1947), pp. 26-28, and K. Ziegler, (Stuttgart, 1949), 36. If Demetrius did hear a Celtic tale of a god or hero asleep on some western island, it would have been easy for him or Plutarch to identify the subject with Cronus (cf. Hesiod, , 169, and Pindar, , ii. 77 [70] ff.; see also note a on p. 182 and note a on 942 A s.v.). Pohlenz’s notion (R.E. xi. 2013) that Posidonius, who was familiar with the northern world, was the intermediary of this Kyffhäusermotiv has nothing to support it. Posidonius was the source of the Cronus-motif as well as of the whole geographical part of the myth according to M. Adler, Op. cit. (note b, p. 18), pp. 169-170, who has no trouble in showing that Schmertosch adduced no real reason for designating Xenocrates as Plutarch’s source for this section; but Hamilton (loc. cit. [note a, p. 22]) has proved that Posidonius could not have been the source either.

The discussion of the second part of the myth, the demonology and eschatology, has also been concerned mainly with the problem of Plutarch’s sources. Heinze attempted to prove that this myth had been put together out of material drawn from Xenocrates and from Posidonius and that in the resulting combination the parts that belong to those two authors are distinguishable.Richard Heinze, (Leipzig, 1892), pp. 123 ff. M. Pohlenz, (Göttingen, 1909), p. 133, n. 1, approved Heinze’s conclusion in general but differed with him in some particulars. Adler vigorously attacked this thesis and argued that Posidonius was Plutarch’s source for the whole myth and for whatever there is in it that may have come ultimately from XenocratesMaximilian Adler, Op. cit. (note b, p. 18), pp. 166 ff. Adler’s dissertation was reviewed by Pohlenz in xxxii (1912), 648-654, and his thesis concerning the source of the myth criticized, ibid. 653. P. Capelle ( [Halle, 1917], pp. 14-15) held that chapter 28 came from Posidonius’s account of the state of souls after death and chapters 29 and 30 from his supposed commentary on the Timaeus.; but R. M. Jones (Chicago Dissertation, Menasha, Wisconsin, 1916), pp. 48-56 and 58-60. proved conclusively that neither Heinze’s conclusions nor Adler’s will bear scrutiny, that Posidonius could not have been the source, and that, while Plutarch combined various eschatological notions which were current and some of which were probably held in common by different philosophers, his myth is in the main an interpretation of Plato’s Timaeus. Later, against Karl Reinhardt’s attempt to trace the myth back to a hypothetical solar esehatology of Posidonius, Jones re-established the Platonic character of Plutarch’s eschatology, psychology, and demonology here and the impossibility of taking Posidonius for the source.K. Reinhardt, (Munich, 1926), pp. 313 ff. (cf. also F. Cumont, La ThÈologie solaire du paganisme romain, , xii [1909]); R. M. Jones, Posidonius and Solar Eschatology, xxvii (1932), pp. 113-135, especially pp. 116-131. P. BoyancÈ, (Bordeaux and Paris, 1936), pp. 78-104, follows Jones in refuting Cumont and Reinhardt. Hamilton later contended even more positively that Plutarch took the as the model for the whole of his myth in the and that, since the shows that he regarded the seriously, he must have intended the corresponding portion of his myth in the to contain an equally serious exposition of his own beliefs concerning the nature and fate of the soul.W. Hamilton, xxviii (1934), pp. 24-30. Hamilton expressly opposes the theory of von Arnim, who, in his Plutarch über Dämonen und Mystik (see note a, p. 22), pp. 24-65, argues that Plutarch took the geographical myth and the eschatological myth from two different sources and the latter from an eclectic Platonist later than Antiochus. As to Hamilton’s notion of the seriousness with which Plutarch intended the myth, Ziegler is surely right in saying (, 217) that Sulla’s final sentence, taken together with Lamprias’s remark in 920 B - C, shows that Plutarch had no intention of insisting upon the literal truth of the myth; in this attitude also he follows Plato: see note a on p. 223 s.v.. Soury in his extensive study of the myth, while emphasizing the possible influence of the mysteries, agrees in general with Hamilton that it is preponderantly Platonic.G. Soury, liii (1940), pp. 51-58, and (Paris, 1942), pp. 73-82 and 177-210.

Anyone who without a preconceived thesis to defend reads the De Facie will recognize, I believe, that Plato was Plutarch’s inspiration throughout the dialogue but that Plutarch is himself the true author of the whole work and that, while there is in it a distillation of his wide and varied scientific and philosophical reading, he cannot possibly have composed it by copying out any source or combination of sources. I have tried in the exegetical notes to indicate the parallels which will help the reader to understand the dialogue itself by seeing its relation to the rest of ancient scientific and philosophical thought. Among these parallels some of the most striking are drawn from later writers, especially Neo-Platonists; these I have mentioned not in order to insinuate that they show Plutarch’s direct influence upon those later writers, although many of them certainly were acquainted with him, but because they illuminate the meaning of the and at the same time indicate what may have been contained in some of the philosophical writings known to Plutarch and long since lost to us, and may help to cast some flicker of light upon that obscure and controversial problem, the prehistory of Neo-Platonism.

6. The , which is No. 73 in the so-called Catalogue of Lamprias and No. 71 in the Planudean order, is apparently preserved in only two MSS. of the Bibliothéque Nationale, Grec 1672 (saec. XIV) and 1675 (saec. XV), conventionally called Parisinus E and Parisinus B respectively.On the MSS. of Plutarch generally cf. the references cited by M. Pohlenz, , i (Teubner, 1925), Praefatio, p. vi, n. 1, and pp. xxvi and xxviii f. on B and E respectively. These have hitherto been supposed to be independent copies of a single archetypeWyttenbach ( [Oxford, 1795], p. xliv) says of B ut videtur, ex E, aut ejusdem exempli codice, ita descriptus ut antiquiores melioresque simul adhiberentur; unde quaedam lacunae uberius etiam expletae, et plura menda sanata. M. Treu, , ii (Ohlau, 1881), pp. 5-7, argued that B derives from the same source as E, which B must have used later; and his conclusion was generally accepted by later editors. Raingeard’s more complicated stemma (p. xiv of his edition of the ) is, in any case, entirely unjustified.; but recently G. R. Manton has put forward strong arguments for thinking that B is a descendant of E through an intermediate manuscript, a copy of E, which was worked over by a scholar who filled in lacunae and inserted conjectures of his own. The Manuscript Tradition of Plutarch Moralia 70-7, xliii (1949), pp. 97-104. Among the passages discussed by Manton where B has readings other than those of E are none from the , for the text of which Manton (Op. cit. p. 99, n. 1) depended upon Treu’s collation supplemented by Bernardakis’ list in vol. i of his edition, pp. 1 ff.; but I have found no variant reading of B in this essay that would surely gainsay Manton’s hypothesis. Those which might suggest that B is not descended from E are the following: 927 F: τὸν -B for E’s correct τὰ before ἐμβριθῆ; 929 B: ἔχων δὲ -B, ἔχων δὲ τοῦτο -B for the correct ἑκὼν δὲ; 932 D: πεποιημένων -B for E’s correct πεπεισμένων: 937 F: ἐπιφερομένη -B, φερομένη -E for the probable original ἀντιφερομένη; 938 D: ἀναγινώσκων -B for E’s correct ἀναγινώσκοντος; 943 D: καταγινομένας -B for E’s correct καταδυομένας, Manton’s conclusion has been rejected by K. Hubert ( xciii [1950], pp. 330-336), but Hubert’s defence of the independence of B and E has been counterattacked by Einarson and De Lacy ( xlvi [1951], pp. 103 and 106, with notes 36 and 56).

I have collated both manuscripts from Photostats which were generously put at my disposal by Dr. William C. Helmbold; and I have recorded under the usual symbols the variant readings of each of them, for I soon discovered that not only is Bernardakis’ report of the mss. untrustworthy, but that the same must be said of Raingeard’s in his recent edition of the dialogue, and that even Treu’s collation (see note b, p. 26) is not free of errors. I have not recorded mere omissions or variations of accent or breathing, however, unless the sense is affected by them; and I have regularized crasis and elision without regard to the manuscripts or report of them, for they show no consistency in this matter.For example, in 931 D they have τὰ αὐτὰ πάσχειν ὑπὸ τοῦ αὐτοῦ ταὐτὰ (B, ταυτὰ -E) ποιεῖν ταὐτὸν . . . and occasionally οἶδ ὅρως and ἀλλ ὅρως, although they do not ordinarily elide the α of oἶda and ἀλλά. Almost invariably both E and B have μὴ δὲ instead of μηδὲ or μηδ. On these matters cf. T. Doehner, (Meissen, 1858 and 1862), especially iii, p. 51, and ii, p. 35, n.%2%2; and on the question of hiatus cf. Helmbold, xxxiii (1938), pp. 244-245, and xlv (1949), pp. 64 f. with his references, and for a much stricter view Ziegler, , 295-298. To emend for the sole purpose of eliminating hiatus is to take unwarranted liberty with the text; but, on the other hand, to introduce hiatus by emendation is certainly inadmissible. It should be observed, however, that in the , besides the exceptions to avoidance of hiatus listed by Ziegler (Op. cit. 296-297), final αι, οι, ει, and ου before an initial vowel may always be possible (cf. for ου e.g. τοῦ ἰδίου ἀέρος in 944 A), ἄνω and κάτω are permissible before any word beginning with a vowel (cf. ἄνω ἔχειν and κάτω ἄνυθεν in 924 C which guarantee ἄνω ἐστιν in 926 A), and other cases of hiatus which cannot reasonably be eliminated occasionally occur (e.g. χείλη εἰκόνας [921 C], τουτὶ εἴρω [935 D]). In conformity with the usage of Professor Babbitt and regardless of the manuscripts, I have printed the forms γίγνεσθαι, γιγνώσκειν, and οὐδείς, though the manuscripts usually have γίνεσθαι, γινώσκειν, and οὐθείς; but I have adopted the form duei=n throughout. I have tried to the best of my ability to assign emendations to those who first proposed them; but for some which appear without ascription in all modern editions, and the author of which I have been unable to discover, I have had to be content with the unsatisfactory note, editors. For the suggestions said to be written in three different hands on the margins of the copy of the Aldine edition now in the Bibliothéque Nationale (RÈs. J. 94), I have had to rely upon the report of Raingeard in the apparatus criticus of his edition (cf pp. xvi f. of his Introduction)P. Raingeard, (Paris, 1935). Raingeard’s text is fantastically conservative, reproducing E for the most part even where E gives impossible Greek; and yet his report of the manuscripts is frequently erroneous either explicitly or by implication. The translation is worse even than the text; and the commentary, especially where it touches upon philosophical and scientific questions, is more often wrong than right, almost everywhere inadequate, and frequently absurd.; all of these I indicate without differentiation by the formula, Anon., Aldine, R.J. 94. Upon Raingeard’s report and those of Reiske, Wyttenbach, Hutten, and Bernardakis I have had to rely for the variant readings of the Aldine edition and of the edition of Xylander; but the edition of Froben (Basiliensis, 1542), as well as those of Stephanus (1624), Reiske, Wyttenbach, Hutten, Dübner, Bernardakis, and Raingeard, and the translations of Xylander, Amyot, Kepler, Kaltwasser, the two translations of Prickard,See note b, p. 18. Prickard’s translation of 1911 was reviewed by W. R. Paton, xxvi (1912), p. 269, and by L. C. Purser, , xvi (1911), pp. 309-324, whose review is rather a series of notes and suggestions for almost two score passages in the essay. and that of portions of the essay by Heath,Sir Thomas L. Heath, (London, 1932), pp. 166-180. I have consulted and compared throughout.

Those emendations which, so far as I know, are original with me are indicated by the initials H. C. Besides the editions, translations, and articles already mentioned in this Introduction, the chief aids to my study of the text have been the following: